Empathy is essential for healthy socialization, being correlated with anonymous donation, spontaneous sharing, sympathetic reactions to distress, and other forms of prosocial behavior. Interventions increasing empathy may be particularly useful for individuals who have become desensitized by violent environments. These individuals' empathic deficits affect their social cognition and decision-making, impeding their social reintegration. A key issue in empathy research is to elucidate how to improve empathic tendencies to facilitate such reintegration. Restoring empathy and prosocial behavior in individuals with low empathy is of high significance for mental health in our society. To do that, we need functional and neurobiological models of empathy that can adopt viable markers, suggest doable interventions and make clear-cut predictions about outcomes. This project proposes the testing of such a model. This project proposes a dual level model of empathy with a core low level supported by neural processes that internally simulate what another person is doing and experiencing, and a cognitive higher level for deliberation of empathic decision-making. It aims first at establishing correlations between low level empathy and high level empathy using functional magnetic resonance imaging (fMRI). It also aims at examining the effect of an emotion imitation intervention on low and high level empathy. Finally, it aims at showing that by disrupting prefrontal activity with rTMS it is possible to release inhibition of imitation, thus favoring low level empathy, which in turn would increase high level empathy.